Studyguide Detailed for First Exam
Studyguide Detailed for First Exam STC 103
Popular in Statistical Reasoning for Strategic Communication
Popular in Communications
This 8 page Study Guide was uploaded by Natalie Land on Tuesday September 13, 2016. The Study Guide belongs to STC 103 at University of Miami taught by Bo Ra Yook in Fall 2016. Since its upload, it has received 9 views. For similar materials see Statistical Reasoning for Strategic Communication in Communications at University of Miami.
Reviews for Studyguide Detailed for First Exam
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 09/13/16
STC 103 Statistical Reasoning • What is research? o Systematic investigation of phenomena that leads us to understand and predict outcomes § Academic research: guided by a theory, select methodology, how will they analyze data, academic research always tests hypothesis: results plus discussions § Begin with a problem question or purpose • How many people are attending class? • What’s a popular class? § Relational: something’s are related but not necessarily X causes Y § Casual: X does cause Y § Research question: statement of the problem • What is the most popular auto brand? (descriptive) no relation of X and Y • Females are more likely to purchase white car than male (relational) § Hypothesis: testable prediction, based on theory or observation (relational) § Research method: strategy plan and activity to accomplish research • How to test hypothesis • Quantitative: objective, systematic, controlled, o Can be generalized, uses numbers and discrete units o Descriptive: seeing things as they are, no manipulation, made of surveys or polls o Experimental: manipulation to see if something different occurs o Why quantitative rather than qualitative: because social science we want to replicate in different target audience § Apply findings to different studies o Choose method based on research objective and what you want to know o UNDERSTAND ADVANTAGES VS DISADVANTAGES o Phenomena: any object or event explained how will you explain the event or whats going on • Qualitative: specific to case, each case is unique, cannot be generalized o Ie : freshman are more international students from Europe and how those feel are different from Africa • Variable: observable characteristic o Independent variable: the variable that represents the cause of the dependent variable § Naturally occurring § Age, biological sex § What we manipulate § Color credibility § Iv causes an effect (DV) o DV: outcome, the effect of the independent variable § Perceived reputation § Variable the research tries to explain • Behavioral: likes, retweets, number of calls your getting • Attidunal: brand preference, trust, relationship, reputation o Iv causes the DV o IV will cause the DV the effect o When people like the brand (IV) the more likely they are to purchase (DV) o • What is Data? o Data: reports of observation of variables o Data reflect something measured § It has no meaning until researchers create a meaningful system for interpretation § Has to be measured o Has to be reliable and valid § Reliable: same results over time: consistent § Valid: measuring what we want to measure, describing what we think it is • Statistics: a branch of applied math that specializes in procedures for describing and reasoning from measures o A lot of procedures each with mathematical deductions o What is a measurement? Links observation to numbers o Data: plural: datum: single observation o Social science: 1 to 5 scale or 1 to 10 scale: very reliable and valid results • Physical observation: such as objective amounts o Advertising: cost per thousand, placement on a page or impressions • Psychological: cant be seen : values, beliefs, attitudes • Concepts that we don’t know how to calculate: so we come up with measurements • Stake hold: is part of the company in any way, employee, etc • Stock holder: owns a stock or share of company • ROI: return on investment: money value but also expectation • Sampling o Large numbers (public, audience) that are too large to get data on so we use sample of entire population § We want to know the entire communication school students: this is population § Sample: just one class o Sample: subset of population • Descriptive statistics: the overall population • Sampling: calculated values that represent how sample characteristics vary from population characteristics • Probabilistic sampling statistics: we give entire population same opportunity to be part of the sample o All names in one bowl • Statistical analysis: decide before hand what procedure used and what criteria involved in reasoning o Social science accepts only five percent error aka ninety five percent confidence in outcome • Measurement o Scheme for assignment of numbers or symbols to specify different characteristics of a variable § It’s the how § Links observation to the number § Standardized o Physical observation: such as objective amounts o Psychological: observations that can’t be seen § Perceptions, influence someone to buy something o Role: bridge of what’s our there in the real world vs interpret that into our study § Observation plus statistical models • Two classifications o Continuous: time spending, how likely o Categorical: gender, years in college, category information • Construct: abstract ideas things we can’t view • Conceptual: verbal meaning of the concept • Operational: translates into prescription for measurement • How to measure variables o Using scales, scale: specific scheme for assigning numbers or symbols to designate characteristic of variables o Scale example: A+ A-‐ these are for grades o Behavioral intention measure: something you want to plan or do § How likely, 1, 2,3,4,5: strongly disagree to agree • Levels of measurement o Different measurement levels offer varying degrees of exactness in describing given characteristics • Two types of variables o Categorical: nominal or ordinal o Continuous: amount of time or money etc § Interval, or ratio o Categorical: nominal § Assignment of numbers to categories into math meaning • Where do you live? • (1) on campus • (2) off campus • the numbers don’t mean anything o ordinal § ordered relations, unspecified intervals, rank ordering: better, faster, stronger § IE: best selling book #1, #2 etc o Interval: to identify ordered relations of characteristics, equal intervals, o Ratio scale: equal distance but has an absolute zero point § Age, distance, time o Categorical: places observation into classes § No values § Reprenstation as counts or percentages § Nominal and ordinal • Nominal: numbers don’t matter • Ordinal: numbers do mater § Set, race, major • Continuous: places observation on a continuum o Mean, median, mode, range o Interval, scale, tatio o Age, income, temperature • Continuous can be categorical but categorical cannot be continuous • Categorical: nominal • Categorical: ordinal • Independent variable: live on campus or not • Dependent Variable: Likely to purchase a car August 31, 2016 • Interval: likely: behavior, thinking process, no absolute zero • Ratio: absolute zero • Measurement level is nominal ordinal ratio or interval • Roadmap to measuring behavior o Behavior of consumer o We measure, values, beliefs, attitudes, opinions: to see the behavior • Attitude measurement o Interval data o Respondents react to statements by degree of agreement § Must have midpoint § 1 to 7, 1 to 5, 1 to 9 § add number of responses § two or more statemenets o Sematic Diffential/ Bipolar Scale § Each side has opposite meaning § Good bad § Easy difficult § Helpful unhelpful • Measurement adequacy o Reliability: internal/ external consistence of measurement o Externally: with the same conditions would be the same result? o Internally: subparts of measurement related to eachother o Temporal stability: stable result over time § Ie: temperature of a fridge o Internal consistency: set of scale gives you the same results § Likelihood to purchase § All give you the same number of more likely to less likely o Four items related to eachother o External: degree to which scroes on measure are stable over time § Test retest correlation: asseses strength of relationship between the same groups score on same measure at two or more points • Overtime very simiiar results at same time § Internal consistency: degree to which the statements all tap on the same thing § Cronback alpha: asseses internal consisteny of the items • From 0 to 1: .8 is most desirable o To increase internal consistency § Avoid item scales • Entire measure unreliable • Adequate number of similiarly worded items § Reliability: same result over time § Validity: something measured that measures what you want to measure • Ie if you measure temperature or want to but instead get the heart rate its not valid § To be useful must be both reliable and valid § Validity implies reliability but reliability not necessarily validity. Chapter 2 STC 103 Notes • Descriptive Statistics: to describe, summarize, and organize o Just explaining what’s going on o Sampling: studying and testing prediction and hypothesis § Experimental o Categorical data: either or classification of categories o Continuous data: continues, could be infinite § Nominal, • Gender numbers don’t mean anything § ordinal, • order matters § interval ratio • how likely • no absolute zero § ratio • absolute zero • Frequency: number of times something happens o Ie : in one week I eat four times at subway o To see the frequency distribution, we have an organized table, table or graph, shows the categories, and the frequency: (number of individuals in each category) § Valid percentage of all valid cases, if a variable is missing not valid • Bar Chart o X axis: has the categories o Y axis: frequencies for each category • Pie Chart: o All together is one hundred percent • Continuous data descriptions o Frequencies: the normal curve o Measure of central tendency: how data is spread or concentrated o Point is plotted above each score or measurement and then you connect the dots § Histogram: horizontal line at each point, corresponds to interval of scores § Frequency polygon: graph representing the frequency of scores in smooth curve points connected • In histogram one score connects to another in frequency it doesn’t • Bar chart is for categorical data and histograms is for continuous data • Continuous Data o How graph will shape? § Symmetrical: data on each side is mirrored § Skewed scores: pile up on one side and tapper off on other § Skewedness: measure of asymmetry if tail is on right its high scores this equals positive skew § § positive skew § if tail is on opposite side it’s a negative skew and it means low scores § kurtosis: height of the curve: the peak • leptokurtic: less difference higher peak • platykurtic: more difference, flatter peak • Central Tendency: o How together the data is around the middle o How scores are clustered § Spread out or together o 3 Ms § mean § mode § median o purpose: find score, most typical or best representative of the entire group o mean: average, sum of scores divided by number of sample o M : sample mean o The u: population mean o Median: middle point middle score o If n is odd identify middle score and if n is even you average the middle pair to find median o What if there is a lot of N scores § if N is odd • median equals N+1 divided by two • if N is even o N/2 plus N+2/2 o Mode: most frequent score o If mean is bigger than median, then we get a positive skewed o If mean is less than median: negatively skewed o Mean is good for the sum of all individuals Ns or when you know value of every score § Not good for extreme scores, ordinal data, nominal or skewed distribution o Median is good for § Skewed distribution, undetermined values, open ended, ordinal data, § Not good for nominal • Measures of dispersion o Variability of scores o Range: highest score minus the lowest score § Based on only two scores o Variability: distance of spread of scores or distance of a score from the mean § Purpose: to describe distribution o Most important measure is variance and standard deviation § Standard deviation: how far individual score from mean § Describes if scores are clustered closely around mean or scattered § Variance is used for population and sample is S squared used for sample o Variance is the sum of all the (individual scores – the average squared) divided by the number of individual numbers n-‐1 o And for the standard deviation its just the square root of the variance • Sample vs Population o Population: the universe of objects § Public, target audience, too large to measure o Sample: portion of that population § Randomly drawn o Sampling: goal is to generalize provide an estimate population o Random probability sampling: each individual in population has equal chance of getting chosen for the sample • Statistic: characteristic of a sample • Parameter: characteristic of a population • Statistical inference: process by which parameter can be estimated • Population distribution: work for starbucks and want to know if UM students attitudes toward starbucks so we measure attitude of all UM students • Sample distribution: draw a sample and ask attitude of only 100 of 5,000 students for example • Sampling distribution: we want to study population but sub group cant explain entire population so we do man samples so many samples of 100 through 5,000 students • Frequency: distribution tied to a particular number of observations and how this number is divided among different categories • Proportion: of total number of unites
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'